Tascam M512 / M520 Upgrades

he figured the draw against 66% of the supply leaving "headroom"...
Click to expand...

Right, but I like headroom, especially when I'm in uncharted territory (with respect to my experience). :o

plus your adding some extra capacetence to the supply
Click to expand...

Actually, I'm not. I'm adding voltage capacity to many of the larger caps, and higher temp/longer-life caps than stock, but everything I selected for replacements was the same capacitance rating. So should I adjust that and pick some caps with a higher capacitance?

as to the decoupling caps i use polyproplene and seems i remember you asking as to the voltage... go with 50V... from rail to rail.. total diff in the rails is between 30-36V so 50's should be fine...
Click to expand...

Okay. Thank you for the suggestion, and the 30-36V from rail to rail makes perfect sense...50V it is for the caps.

Thanks, demented. ;)
 
sweetbeats said:
Ohhhhhkay...cool. Thank you.



I assume you mean opa2134 right?

Hmmm...so now I need to figure out what the 15V rails are powering...I'm assuming it is not just channel cards but monitor amp stuff and more...So it sounds like there is a possibility the PS will have to be modded if I went with all opa2134's since that would be 120 opamps just for the cannel cards, or select some lower draw opamps for less critical paths...
Click to expand...

No actually the opa2132. The 2132 is a 2134 with poorer specs (whether they are the same die that they bin out or a slightly different process I could not tell you.

I use the 2132 in most all my opamp replacements. Sometimes I use the opa2227 when I have them around. But for the most part they are interchangable.

As for current draw you need to factor in that there are more than one IC type in the mixer and that tl072 are the least power hungry of those that you could be replacing. Also, the regulator IC are rated at 1.5 amp 100% duty cycle and can do much more current for a shorter period of time.

The question to ask is: What am I trying yo get with this mod?

If you wanted to reduce the hiss then you would need to locate the source. THe likely parts are the gain setting resistors of the input stage....

-Ethan
 
Hey sweetbeats. Just thought I'd give you some food for thought here. Not the same board but I have a M320B that I totally re-capped and did the first eight channels with a combination of LM4562 and OPA2134. I kinda ran into the same issue of not wanting to overload the power supply so I never did the other channels with the new op-amps. Couldn't find enough info on how to beef up the supply or build an external which is what I'd love to do if I knew how. Anyway I'm not so sure the opamp upgrade is really worth it. I know it wasn't cheap. I bought enough opamps to do the whole board and will probably end up selling the rest I have cause I don't want to take a chance on burning up the power supply. Actually I hear very little difference in the sound between the channels. There is a difference but very minimal improvement if you ask me. This board is very neutral and clear through all the channels. I would recommend doing all the caps though, but if you leary about the power supply I would think twice about doing the opamps. Maybe just do a pair of channels first and see if you like them.

Oh if I were you I'd re-cap the power supply first to see if it quiets things down a bit.
 
dementedchord said:
i personally cant verify this... if i were you i'ld probably put a .1mf cap from pin 4-8... or at the very least check it with a good oscilloscope after doing the work cause if it's oscillating you aint gonna hear it...
Click to expand...

Shooting from a week back, but:

Standard modern nomenclature for microfarad is uF, not mF. Technically it's really µF, but typing that mu character can be a pain in the ass, so . . .

"mF" means millifarad, which is 1000 times a microfarad.

My personal favorite is HVAC guys, who still refer to microfarads as MFD, which really means megafarad, or 10^12 times microfarad :D . . . you see that in old datasheets sometimes too.


Also, I read a recent analysis of capacitor distortion, which demonstrated no reduction in electrolytic distortion when bypassed with a small-value poly cap. I haven't tested that myself since I read it, and I always used to do it, but just some food for thought. I'll have to see if I can find that link . . .
 
No actually the opa2132. The 2132 is a 2134 with poorer specs
Click to expand...

What's the point of going with the 2132? Is it a cost-based decision?

The question to ask is: What am I trying yo get with this mod?

If you wanted to reduce the hiss then you would need to locate the source. THe likely parts are the gain setting resistors of the input stage...
Click to expand...

The opamp mod is for the purpose of increasing headroom and improving transient response, although in my very limited work with the M-520 thus far, I've been pleased with the sound as-is, using a passive electric bass into one of the onboard hi-Z inputs)...for my ears the e-bass seems to be a good indicator. This isn't to say that my ears are tuned to really great gear :rolleyes:, but compared to that with which I have worked the M-520 treated my bass nicely.

Hiss is another issue, but I though old caps were often the greatest contributor to self-noise in aged gear...as you know I'm going to recap over time, starting with the PS first. The gain setting resistors to which you refer...are you talking about the TRIM pots?

Hey sweetbeats. Just thought I'd give you some food for thought here. Not the same board but I have a M320B that I totally re-capped and did the first eight channels with a combination of LM4562 and OPA2134. I kinda ran into the same issue of not wanting to overload the power supply so I never did the other channels with the new op-amps. Couldn't find enough info on how to beef up the supply or build an external which is what I'd love to do if I knew how. Anyway I'm not so sure the opamp upgrade is really worth it. I know it wasn't cheap. I bought enough opamps to do the whole board and will probably end up selling the rest I have cause I don't want to take a chance on burning up the power supply. Actually I hear very little difference in the sound between the channels. There is a difference but very minimal improvement if you ask me. This board is very neutral and clear through all the channels. I would recommend doing all the caps though, but if you leary about the power supply I would think twice about doing the opamps. Maybe just do a pair of channels first and see if you like them.

Oh if I were you I'd re-cap the power supply first to see if it quiets things down a bit.
Click to expand...

Daveg62, thanks for the perspective. Very valuable to get a real-world impression from somebody who has "been there". ;)
 
thnx for the reminder #1... and yes i'ld like to see that info on the by-pass caps... SBs my bad on the extra supply caps i seemed to remember we talked about an upgrade at some point...
 
Lt1358

Take a look at the opamp thread (I'm sure you have), download the paper and read the comments.

So, instead of using the 2132/34 (I use them interchangeably in most things) you might think about using the LT1358. This has a lower current need than the opa2134 and is a significant upgrade the the TL072 (duh, almost anything is). I would expect that you could use them everywhere in the m520 (except where drive currents greater than 25mA are needed - need to see if there are anyplaces like that). No PS redesign needed.

Plus as you may know I was planning on using them in the MS16 channel cards.

Now to find the best price on them ($4 a device).....
 
SBs my bad on the extra supply caps i seemed to remember we talked about an upgrade at some point...
Click to expand...

demented...yeah I think it did come up somewhere...It just started getting hairy as far as space and my limited knowledge about what was prudent when I started looking at upping the capacitance values from stock, so I'm sticking with same capacitance values but upping the temp, life and voltage ratings. No bad.
 
Ethan...! It happened again! :eek:

I think you responded while I was composing or something because there again is a post from you I didn't see before...huh!

But is the LT1358 just overkill? I mean would I figuratively be polishing a turd? (I'm using extremist figurative language...I don't consider my M-520 a turd, but I do understand its place in the continuum...Its not a Neve, and its not a Fisher Price...)

I would expect that you could use them everywhere in the m520 (except where drive currents greater than 25mA are needed - need to see if there are anyplaces like that)
Click to expand...

I wish I could figure that out...you can gather that from the schematics I assume? Teach me, Obi Wan...

Now to find the best price on them ($4 a device)...
Click to expand...

And you're finding them for $4 where? Best I've found is $4.27 at a quantity of 100+...that adds up...:eek:
 
sweetbeats said:
By "paper" did you mean the tangentsoft article or the sg-acoustics paper?

OPA227 for $2.20...2134 for $1.46...
Click to expand...

Both papers. The sg gives you one hard core engineers view of what makes a good audio opamp. The tangent gives someones soft feel for what the opamps sound like to them. very soft and subjective.


-Ethan
 
sweetbeats said:
The opamp mod is for the purpose of increasing headroom and improving transient response
Click to expand...

Transient response, yes. Less distortion too. Headroom, only to a small extent. Headroom is defined by the rails of the power supply. The only difference an opamp can make is whether it's rail-to-rail, and maybe how much DC offset it has. So figure if you're going from a chip that only goes to say 1V of its rails to a rail-to-rail chip, you'll pick up less than 1dB of additional headroom.

If you want much more headroom, you'd have to rebuild the power supply for a higher voltage--and take into account the effect of that on all the other components in the board . . .
 
Ah yes...faint bells of fractured memory are tinkling(regarding headroom)...I think there may be some headroom gains though especially with the TL072's...IIRC many upgrades could do a better job of taking advantage of the 15V rails...thanks, msh. ;)
 
FOr the input card of your m520

I took a look at the M-520 input card and thought about opamp upgrades.

There are 7 IC listed on the card. We can discard U4 (NJM2903D) right away because it is a comparator used to the peak led and not in the signal path.

U1 TL072CP - this opamp is operated at internel signal levels and is not stressed very much. With that said I note that it forms part of the feedback loop of the mic preamp. I think that if you use the mic pre's you should upgrade this opamp (opa2132 would be great).

U2 NJM4556 - Used for the tape in path - this is a bipolar opamp but I expect that you could use a FET opamp here (such as the opa2132). The NJM4556 is capable of driving 70 mA but that much output drive is not needed in this stage.
Replacement should be considered.

U3 NJM2041DD - a "low noise" bipolar opamp that appears to only be used in the phono circuit. No need to do anything here. If you want outstanding phono buy an outboard phono amp (IMHO)

U4 - NA

U5 TL072CP
U6 TL072CP - these 2 IC's form the EQ section. I would replace them with OPA2132 in a heartbeat.

U7 NJM4556 This IC forms the driver to the mixing busses. THis is where the 70 mA output current may be important. I would have to do much more looking before I could offer a suggested upgrade. It may be that the 40 mA drive of the opa2132 would work. Hard to say without testing.

Looking more closely at the PS I see that the 7815/7915 voltage regulators are used to drive a series pass transistor which does the actual voltage regulation. (2sa1264/2sc3181) These power transistors are rated for 8 amps or more.... Thus the +-15 volt supplys are limited by the power transformer rating (unknown) and the fuse ratings (4 amps). That would be some 500 opa2132 at idle.


So the bottom line is that you could replace all the TL072 with opa2132 without worrying about the power supply. Regarding caps in the PS. Other than using 105 degree caps there is no need to use a higher value on them.

I think that someday I'll do this on my M-520....

--Ethan

PS page 177/178 (M520 PS) lists U4 as 7815 and U5 as 7915. This is a typo. These units form the +-18 volt supply and should be shown as 7818 and 7918 respectivly.
 
ooo...oooo...coool.

The thing about U3 though is that to my eyes and brain U3 is used for the PHONO IN on channels 3 & 4 but then is used for the LINE INs on channels 5 ~ 20. So where does that put things?

U2 NJM4556 - Used for the tape in path - this is a bipolar opamp but I expect that you could use a FET opamp here (such as the opa2132). The NJM4556 is capable of driving 70 mA but that much output drive is not needed in this stage.
Replacement should be considered.
Click to expand...

So if that much drive is not needed, why did they spec the 4556? :confused:

OPA2132 are too spendy for me, I think...what is the big difference between the 2132 and the 2134?

Thanks for all this info, Ethan. This is really helpful.

Also, thanks for the catch on the typo. I've found a couple discrpancies in the parts lists too. Always helpful to confirm visually on the component or double-check the schematic.
 
Oops - RE U3, it is used for the line in. You could put opa2132/34 in there.

opa2132 have better DC specs than opa2134. Nothing that you would or should hear. Where I say 2132 you can read opa2134.

Now if youu wamt to eliminate coupling caps having very low DC offset (good DC specs) would become important.

As to why they speced a 70 mA drive opamp? Who knows? If you are driving low impedence loads or a long buss then having the drive is needed. EQ might be low impedence too so driving it would need the current. The njm4558 could drive about 50 mA (with lots of distortion) ....

-Ethan

sweetbeats said:
ooo...oooo...coool.

The thing about U3 though is that to my eyes and brain U3 is used for the PHONO IN on channels 3 & 4 but then is used for the LINE INs on channels 5 ~ 20. So where does that put things?



So if that much drive is not needed, why did they spec the 4556? :confused:

OPA2132 are too spendy for me, I think...what is the big difference between the 2132 and the 2134?

Thanks for all this info, Ethan. This is really helpful.

Also, thanks for the catch on the typo. I've found a couple discrpancies in the parts lists too. Always helpful to confirm visually on the component or double-check the schematic.
Click to expand...
 
Ethan...Bypass caps...

0.1 or 0.01 uF will do the trick. PS rail to PS rail and/or each rail to ground
Click to expand...

So poly caps in 0.1 or 0.01 uF...this is nuts...Digikey doesn't have wnything in a poly cap at those capacitance ratings...do I just go with anything in between? And when you said "and/or" does that mean that I need to go rail to rail and rail to ground, or did you mean to say "either/or"?

U7 NJM4556 This IC forms the driver to the mixing busses. THis is where the 70 mA output current may be important. I would have to do much more looking before I could offer a suggested upgrade. It may be that the 40 mA drive of the opa2132 would work. Hard to say without testing.
Click to expand...

Testing...you mean like putting test signals into the channel, assigning to a mix buss and scoping the output at the buss out jack for level and distortion? I'm fishing here...
 
Ceramic disc will work fine for bypass caps - the value is not too important at that most oscillations happen in the MHz range not in the audio range.

A quick low cost bypass scheme is to put 1 cap between the + rail and the - rail pins at the socket. The next step up is to go from the + rail to ground and the - rail to ground. Then the next step up is to do all 3 (rails to ground and rails to each other.

sweetbeats said:
Ethan...Bypass caps...

Testing - something like that. Or you could locate or put in a shunt resistor (a resistor in series with the output) and measure the voltage drop across it. Ohm's law (E/R=I) says that the voltage drop E across the resister R gives us the current flow. Drive the hell out of the board to see the peak voltage drop....

Regards, Ethan
Click to expand...
 
A quick low cost bypass scheme is to put 1 cap between the + rail and the - rail pins at the socket. The next step up is to go from the + rail to ground and the - rail to ground. Then the next step up is to do all 3 (rails to ground and rails to each other.
Click to expand...

The rails to ground...so can those just be installed like the rail-to-rail caps? i.e. underneath from pin to pin of the IC?
 
You must log in or register to reply here.
Back
Top